Robert Murphy Someone entirely different (sorry Bob) says in the comments
Homeownership rates are still well above historical norms. We were way overexpanded, even if that doesn’t show up in the aggregates you are using. You do keep emphasizing your point, but I don’t see how it can adequately explain the behavior of ownership rates.
Right but homes and homeownership are not the same. Many people rent homes. Presumably people with worse credit are more likely to rent and a decline in credit standards would move some renters to buying.
However, the distribution of homeownership is different than the number of homes per person. For example, even here in humble Raleigh we saw apartments selling their units during the boom as condos. This would drive up homeownership but do nothing to create more homes.
I would presume we also saw fewer people with home trails. I don’t know the standard term but in the past you saw big savers buy a starter home, keep saving for a down payment and then buy a second home and rent the starter. Then buy a third home later and rent the first two.
As housing prices rose, the incentive to sell an old home rose and the difficulty of putting together a down payment for a new home rose, so I would assume these home trails shortened.
So these are the basic stories where you could see a difference. I don’t know what the data on them are but the core point is that thinking about how many people own homes is separate question from how many homes there are per person.
One major change over time has been a decline in the fraction of new home units that are in structures with 5 units are more. Presumably the 5 unit or more structures are more likely to be rented. So this suggests a much smaller fraction of rentals coming online.